Open guynir42 opened 9 months ago
Maybe add a bunch of columns to the source list table right when we calculate it.
This may even be a shortcut to allow doing astro_cal and photo_cal without having to re-fetch Gaia data.
In calc_aper_cor
in SourceList must come back to using some sort of star classifier, but not necessarily the is_star property.
Confirmed that this indeed fails to find sufficient stars when calculating aperture corrections on the sample of PTF images I use in the fixtures (as of PR #155 ).
Probably best thing to do is to cross match to Gaia (or additional catalogs) at the extraction phase, and then use one of those catalogs to tell if a source is a star, a galaxy, or something unknown (e.g., a supernova).
This will then be propagated through the pipeline and help choose the stars to use in PSF finding or in ZP estimation.
We will probably need to also check if there is already the correct catalog matched to the sources when doing astrometric/photometric calibration and only do a cross match to the required catalog if that hasn't happened yet.
Make sure to integrate the new star/galaxy catalog from Adam Miller's group.
There's a
_is_star
property of the source list that is lazy loaded from the data table'sCLASS_STAR
but it is not very reliable. Right now we are mostly depending on the catalog itself to provide only stars and then the photometric/astrometric calibration can go on without usingis_star
but maybe we can do better by saving the information from the catalog into the source list?Right now the data doesn't live in the database but in the fits table, so I don't know how we would add this.